Project code added

ashutosh · ashutosh · commit f71acf22cc07 · 2025-05-15T12:53:36.000+05:30
diff --git a/README.md b/README.md
@@ -1 +1,9 @@
-# doc-ai-bot
+"""
+    ## AI-Powered Document Q&A Bot
+   ## Upload a PDF and ask questions. Uses LangChain, FAISS, and OpenAI.
+
+    ## Run Instructions:
+    1. Install: `pip install -r requirements.txt`
+    2. Run API: `uvicorn app.main:app --reload`
+    3. Run Streamlit: `streamlit run streamlit_ui/app.py`
+"""
diff --git a/app/llm_pipeline.py b/app/llm_pipeline.py
@@ -0,0 +1,11 @@
+from langchain.chains.question_answering import load_qa_chain
+from langchain.llms import OpenAI
+from vector_store import search_similar
+
+llm = OpenAI(temperature=0)
+chain = load_qa_chain(llm, chain_type="stuff")
+
+
+def query_answer(query: str):
+    docs = search_similar(query)
+    return chain.run(input_documents=docs, question=query)
diff --git a/app/main.py b/app/main.py
@@ -0,0 +1,17 @@
+from fastapi import FastAPI, UploadFile
+from llm_pipeline import query_answer
+from vector_store import ingest_document
+
+app = FastAPI()
+
+
+@app.post("/upload")
+async def upload_pdf(file: UploadFile):
+    content = await file.read()
+    ingest_document(content)
+    return {"status": "Document ingested."}
+
+
+@app.get("/query")
+def query(q: str):
+    return {"answer": query_answer(q)}
diff --git a/app/utils.py b/app/utils.py
diff --git a/app/vector_store.py b/app/vector_store.py
@@ -0,0 +1,23 @@
+from langchain.vectorstores import FAISS
+from langchain.embeddings.openai import OpenAIEmbeddings
+from langchain.text_splitter import CharacterTextSplitter
+from langchain.document_loaders import PyPDFLoader
+import tempfile
+
+embedding = OpenAIEmbeddings()
+vector_db = None
+
+
+def ingest_document(content: bytes):
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp:
+        tmp.write(content)
+        loader = PyPDFLoader(tmp.name)
+        docs = loader.load()
+        splitter = CharacterTextSplitter(chunk_size=500, chunk_overlap=50)
+        chunks = splitter.split_documents(docs)
+        global vector_db
+        vector_db = FAISS.from_documents(chunks, embedding)
+
+
+def search_similar(query: str):
+    return vector_db.similarity_search(query, k=3) if vector_db else []
diff --git a/requirements.txt b/requirements.txt
@@ -0,0 +1,8 @@
+langchain
+openai
+faiss-cpu
+streamlit
+PyMuPDF
+uvicorn
+fastapi
+python-multipart
diff --git a/streamlit_ui/app.py b/streamlit_ui/app.py
@@ -0,0 +1,15 @@
+import streamlit as st
+import requests
+
+st.title("📄 Ask your Document")
+file = st.file_uploader("Upload your PDF")
+if file:
+    response = requests.post(
+        "http://localhost:8000/upload", files={"file": file})
+    st.success("File uploaded")
+
+query = st.text_input("Ask a question")
+if query:
+    answer = requests.get("http://localhost:8000/query",
+                          params={"q": query}).json()
+    st.write("Answer:", answer["answer"])
